*Manuscript Click here to view linked References

1 1 2 3 2 Environmental setting of deep-water oysters in the Bay of Biscay 4 5 6 3 7 8 4 Van Rooij, D.a, De Mol, L.a, Le Guilloux, E.b, Wisshak, M.c, Huvenne, V.A.I.d, Moeremans, 9 10 5 R.e,a & Henriet, J.-P.a 11 12 13 6 14 15 7 a Renard Centre of Marine Geology, Ghent University, Krijgslaan 281 S8, B-9000 Gent, Belgium 16 17 8 b IFREMER, Laboratoire Environnement Profond, BP70, F-29280 Plouzané, France 18 19 9 c GeoZentrum Nordbayern, Erlangen University, Loewenichstr. 28, D-91054 Erlangen, Germany 20 21 10 d Geology and Geophysics Group, National Oceanography Centre, European Way, SO14 3ZH 22 23 11 Southampton, United Kingdom 24 25 12 e Scripps Institution of Oceanography, UCSD, La Jolla, California, United States of America 26 27 13 28 29 30 14 31 32 15 33 34 16 Manuscript for Deep-Sea Research, part I 35 36 17 37 38 18 39 40 41 19 42 43 20 44 45 21 46 47 48 22 49 50 23 * Corresponding author: 51 52 24 Dr. David Van Rooij 53 54 25 Tel.: +32-9-2644583 55 56 26 Fax: +32-9-2644967 57 27 E-mail: [email protected] 58 59 60 61 62 1 63 64 65

1 28 Abstract 2 3 29 We report the northernmost and deepest known occurrence of deep-water pycnodontine 4 5 6 30 oysters, based on two surveys along the French Atlantic continental margin to the La 7 8 31 Chapelle continental slope (2006) and the Guilvinec Canyon (2008). The combined use 9 10 11 32 of multibeam bathymetry, seismic profiling, CTD casts and a remotely operated vehicle 12 13 33 (ROV) made it possible to describe the physical habitat and to assess the oceanographic 14 15 16 34 control for the recently described species zibrowii. These oysters have 17 18 35 been observed in vivo in depths from 540 to 846 m, colonizing overhanging banks or 19 20 21 36 escarpments protruding from steep canyon flanks. Especially in the Bay of Biscay, such 22 23 37 physical habitats may only be observed within canyons, where they are created by both 24 25 38 long-term turbiditic and contouritic processes. Frequent observations of sand ripples on 26 27 28 39 the seabed indicate the presence of a steady, but enhanced bottom current of about 40 29 30 40 cm/s. The occurrence of oysters also coincides with the interface between the Eastern 31 32 33 41 North Atlantic Water and the Mediterranean Outflow Water. A combination of this 34 35 42 water mass mixing, internal tide generation and a strong primary surface productivity 36 37 38 43 may generate an enhanced nutrient flux, which is funnelled through the canyon. When 39 40 44 the ideal environmental conditions are met, up to 100 individuals per m² may be 41 42 45 observed. These deep-water oysters require a vertical habitat, which is often 43 44 45 46 incompatible with the requirements of other sessile organisms, and are only sparsely 46 47 47 distributed along the continental margins. The discovery of these giant oyster banks 48 49 50 48 illustrates the rich biodiversity of deep-sea canyons and their underestimation as true 51 52 49 ecosystem hotspots. 53 54 55 50 56 57 58 59 60 61 62 2 63 64 65

1 51 Keywords: Bay of Biscay, deep-water oysters, canyons, habitat, resuspension, MOW, 2 3 52 Neopycnodonte zibrowii 4 5 6 53 7 8 54 1. Introduction 9 10 11 55 Ocean margins are dynamic environments that host valuable deep-water benthic 12 13 56 ecosystems. Along the Eastern Atlantic margin from Morocco to Norway, several 14 15 16 57 „deep-water ecosystem hotspots‟, associated with a complex interplay of oceanography, 17 18 58 geology and seabed morphology, have been identified (Weaver and Gunn, 2009). One 19 20 21 59 of these hotspots is represented by canyon ecosystems, which often feature cold-water 22 23 60 coral reefs (Arzola et al., 2008; Canals et al., 2006; De Mol et al., in press; de Stigter et 24 25 61 al., 2007; Dorschel et al., 2009; Palanques et al., 2009). The main ecosystem driver 26 27 28 62 within canyons involves a careful balance of the hydrodynamic environment controlling 29 30 63 sediment and nutrient supply (Dorschel et al., 2009; Mienis et al., 2007; Roberts et al., 31 32 33 64 2006). As such, canyons play a critical role since they are the most important 34 35 65 mechanism of focussed nutrient input into the deep marine environment (Canals et al., 36 37 38 66 2006; de Stigter et al., 2007; Duineveld et al., 2001; Palanques et al., 2009). Moreover, 39 40 67 due to frequent incisions during glacial sea-level lowstands (Bourillet et al., 2006; 41 42 68 Toucanne et al., 2009; Zaragosi et al., 2000), the eroded canyon flanks may offer an 43 44 45 69 environment that promotes the settling of sessile organisms, which profit from the 46 47 70 enhanced nutrient flux. Already in the late 19th and the mid-20th century fisheries 48 49 50 71 research had demonstrated the presence of cold-water corals and associated species in 51 52 72 the vicinity of the canyons in the northern Bay of Biscay (Reveillaud et al., 2008). 53 54 55 73 Sporadically, scientists and fishermen also reported dead or sub-fossil oyster specimens 56 57 58 59 60 61 62 3 63 64 65

1 74 from this part of the margin (Le Danois, 1948; Reveillaud et al., 2008), which was not 2 3 75 given the appropriate attention at that time. 4 5 6 76 7 8 77 Although oysters are commonly referred to as typical shallow-water and occasionally 9 10 11 78 reef-forming molluscs, a number of samples have also been recovered from the deeper 12 13 79 realm (Wisshak et al., 2009a). The deep-water oyster, Neopycnodonte zibrowii Gofas, 14 15 16 80 Salas and Taviani 2009 was formally described only recently in Wisshak et al. (2009a) 17 18 81 based on submersible observations and sampling of live specimens in the Azores 19 20 21 82 Archipelago between 2002 and 2007. Further isolated records of this species stem from 22 23 83 steep open slopes such as the Gorringe Bank off Portugal (Auzende et al., 1984), south 24 25 84 of Madeira (Hoernle et al., 2001) and in the Central Mediterranean Sea, where they 26 27 28 85 occur as prominent (sub-) fossil oyster banks (Gofas et al., 2007). This „living fossil‟ 29 30 86 oyster is most unusual with respect to its habitat, size, geochemical signature and its 31 32 33 87 particularly pronounced centennial longevity (Wisshak et al., 2009a, b). It stands in 34 35 88 strong contrast to all other extant oyster species, such as Neopycnodonte cochlear (Poli, 36 37 38 89 1791), which are relatively short-lived. The smaller (4-5 cm) N. cochlear was reported 39 40 90 by Le Danois (1948) on the upper slope of the Bay of Biscay (200-500 m). It is usually 41 42 91 associated with hard substrates and in certain places colonizes Dendrophyllia cornigera 43 44 45 92 (Lamarck, 1816) coral reefs (Le Danois, 1948). Neopycnodonte cochlear certainly has 46 47 93 the largest distribution worldwide, both in ancient (Videt and Neraudeau, 2003) and 48 49 50 94 modern environments (Harry, 1981). The recently described Neopycnodonte zibrowii 51 52 95 can be regarded as a distinct deep-sea relative of N. cochlear with specific adaptations 53 54 55 96 allowing it to thrive in upper bathyal depths. Most recently, Delongueville and Sciallet 56 57 97 (2009), reinvestigated two unusually large specimens sampled alive from the Bay of 58 59 60 61 62 4 63 64 65

1 98 Biscay margin and previously identified as N. cochlear (Delongueville and Sciallet, 2 3 99 1999). These can now be attributed to N. zibrowii. 4 5 6 100 7 8 101 In this paper, we describe the physical and oceanographic setting of the northernmost 9 10 11 102 and deepest occurrence of Neopycnodonte zibrowii oysters within two canyons along 12 13 103 the French Atlantic margin (Fig. 1). Seabed observations with the Remotely Operated 14 15 16 104 Vehicle (ROV) Genesis resulted in the discovery of giant deep-water oyster banks and 17 18 105 cliffs at depths between 540 to 846 m (Table 1). These observations were performed 19 20 21 106 with R/V Belgica during the HERMES Geo cruise in June 2006 (La Chapelle 22 23 107 continental slope) and the BiSCOSYSTEMS cruise in June 2008 (Guilvinec Canyon). 24 25 108 Although no samples could be acquired for morphological or molecular , the 26 27 28 109 observed spatial distribution of the molluscs in relation to the slope morphology, and 29 30 110 local hydrographic regime, provides an insight in the habitat requirements of this 31 32 33 111 recently described species. 34 35 112 36 37 38 113 2. Regional setting 39 40 114 The morphology of the NE Atlantic continental margin in the Bay of Biscay is 41 42 115 characterized by spurs and canyons, organized in drainage basins and feeding deep-sea 43 44 45 116 fans during glacial times (Bourillet et al., 2003; Zaragosi et al., 2000). 46 47 117 48 49 50 118 Most of the water masses in the Bay of Biscay are of North Atlantic origin (Pollard et 51 52 119 al., 1996). The uppermost water mass is the Eastern North Atlantic Water (ENAW), 53 54 55 120 which extends to depths of about 400 to 600 m. Although this water mass has a salinity 56 57 121 of 35.6, according to Pollard et al. (1996), there is a core of low density water around 58 59 60 61 62 5 63 64 65

1 122 500 m water depth, corresponding to the lateral influence of the Subarctic Intermediate 2 3 123 Water (SAIW). Between 400 to 500 m and 1500 m water depth, the Mediterranean 4 5 6 124 Outflow Water (MOW) follows the continental slope as a contour current. Its cyclonic 7 8 125 circulation is conditioned by seafloor irregularities and the Coriolis effect. MOW 9 10 11 126 velocities have been measured in the Bay of Biscay at 8ºW and 6ºW with minimum 12 13 127 values of 2-3 cm/s. Low salinity values observed on the Armorican continental slope 14 15 16 128 may reflect a depletion of the MOW core (Van Aken, 2000). Between 1500 and 3000 m 17 18 129 water depth, the North Atlantic Deep Water (NADW) is recognized. It includes a core 19 20 21 130 of the Labrador Sea Water (LSW) at depths between about 1800 and 2000 m (McCave 22 23 131 et al., 2001; Vangriesheim and Khripounoff, 1990). Below the NADW, the Lower Deep 24 25 132 Water mainly seems to result from the mixing of the deep Antarctic Bottom Water and 26 27 28 133 the Labrador Deep Water (Haynes and Barton, 1990; Van Aken, 2000). 29 30 134 31 32 33 135 Along the slopes of the Bay of Biscay strong, localized internal tides are reported, 34 35 136 resulting from a combination of favourable water mass stratification, steep topography 36 37 38 137 and strong barotropic tidal currents (Huthnance, 1995; Pairaud et al., 2008; Pingree and 39 40 138 Le Cann, 1989, 1990). As the slope is intersected by canyons, these tidally induced 41 42 139 transports may be channelled, creating regions of locally increased bottom flow with 43 44 45 140 mean values of about 14 cm/s or higher (Pingree and Le Cann, 1989, 1990). The action 46 47 141 of upper slope internal tides is proposed to explain the enhanced levels of surface 48 49 50 142 phytoplankton abundance (Holligan et al., 1985; Pingree et al., 1982). 51 52 143 53 54 55 144 3. Material and methods 56 57 58 59 60 61 62 6 63 64 65

1 145 The data collected for this study was acquired during two R/V Belgica expeditions in 2 3 146 the Bay of Biscay. A first campaign, HERMES Geo, focussed during a period of 3 days 4 5 6 147 in June 2006 on the La Chapelle continental slope (Fig. 1b). In June 2008, the 7 8 148 BiSCOSYSTEMS cruise surveyed the vicinity of the Guilvinec Canyon (Fig. 1c). 9 10 11 149 During both cruises a geophysical survey (multibeam bathymetry and seismic profiling) 12 13 150 and CTD profiling (Fig. 2) preceded the ROV observations. 14 15 16 151 17 18 152 3.1 Geophysical survey 19 20 21 153 Initial swath bathymetry coverage of both study areas was obtained using the R/V 22 23 154 Belgica hull-mounted SIMRAD EM-1002 multibeam echosounder. On the La Chapelle 24 25 155 continental slope, an area of 72 km² was mapped at water depths ranging from 200 to 26 27 28 156 950 m (Fig. 1b) and processed using the IFREMER CARAIBES software. In 2008, an 29 30 157 area of 584 km² was mapped around the Guilvinec Canyon at water depths ranging from 31 32 33 158 180 to 1000 m (Fig. 1c). The 2008 dataset was processed using the MB-Systems and 34 35 159 IVS Fledermaus software. Both datasets were gridded to a cell measuring 20 by 20 m 36 37 38 160 and visualized using GMT version 4.2 (Wessel and Smith, 1991). 39 40 161 41 42 162 In order to obtain insights into the sedimentary processes and the thickness of the 43 44 45 163 sediment cover, single channel seismic profiles were acquired using a SIG sparker 46 47 164 source at a velocity of 3 knots (Figs. 1c and 3). The source was triggered every 3 s, 48 49 50 165 reaching 500 J energy. The vertical resolution of the profiles varied between 1 to 2 m. A 51 52 166 basic processing (band-pass filtering, automatic gain control) was applied using the 53 54 55 167 PROMAX software. 56 57 168 58 59 60 61 62 7 63 64 65

1 169 3.2 Water mass characterization 2 3 170 At both sites, information regarding the water mass stratification was obtained using a 4 5 6 171 SEACAT SBE 19 CTD down to water depths of about 1400 m (Table 2, Fig. 2). The 7 8 172 raw data was binned at 1 m using SBE Data Processing (version 7.16a). The obtained 9 10 3 11 173 temperature (°C), salinity and derived potential density (sigma-thèta, kg/m ) were used 12 13 174 to identify the water masses and to indicate the relationship between deep-water oyster 14 15 16 175 occurrence and hydrography (Fig. 2). The zig-zag pattern observed below 800 m in the 17 18 176 salinity and density curves of cast B0813-CTD-4, may be due to a defective CTD pump, 19 20 21 177 and should be considered with caution. 22 23 178 24 25 179 3.3 ROV observations 26 27 28 180 The ROV observations (Table 1, video files as online supplementary material) were 29 30 181 performed using Ghent University's ROV Genesis, a Sub Atlantic Cherokee type ROV 31 32 33 182 with a Tethered Management System (TMS) allowing investigations down to 1400 m 34 35 183 water depth. The underwater positioning was obtained using an IXSEA GAPS USBL 36 37 38 184 system, allowing an accuracy in the order of 2 m. Seafloor observations were made by 39 40 185 means of a forward-looking colour zoom and black & white video camera, assisted by > 41 42 186 250 Watt Q-LED illumination. Laser markers were added to the camera head for scale 43 44 45 187 (10 cm spacing). High-resolution images were acquired at irregular intervals using a 46 47 188 digital Canon Powershot stills camera. Unfortunately, due to a defect during the June 48 49 50 189 2006 campaign, images had to be derived from video capture instead of the stills 51 52 190 camera. The processing and interpretation of dive B06-02 was performed in an ArcGIS 53 54 55 191 environment, expanded with the Adélie extension for ArcGIS 9.0 developed at 56 57 192 IFREMER. Dives B08-02 and B08-05 were interpreted using OFOP (Ocean Floor 58 59 60 61 62 8 63 64 65

1 193 Observation Protocol) version 3.2.0c (Huetten and Greinert, 2008) and integrated into 2 3 194 ArcGIS. For the purpose of this paper, the ROV data were only interpreted with respect 4 5 6 195 to the occurrence of Neopycnodonte zibrowii and the different substrates. A more 7 8 196 detailed study on the distribution, diversity and habitat settings of cold-water corals 9 10 11 197 within the Guilvinec Canyon was performed by De Mol et al. (in press). 12 13 198 14 15 16 199 4. Results 17 18 200 4.1 Environmental setting 19 20 21 201 4.1.1 Hydrography 22 23 202 The CTD casts show a similar water mass stratification in both study areas (Fig. 2). The 24 25 203 seasonal thermocline is recognized down to 50 m water depth (Figs. 2b-c). A salinity 26 27 28 204 minimum (± 35.58) at about 550 m separates the upper Eastern North Atlantic Water 29 30 205 (ENAW) from the saline Mediterranean Outflow Water (MOW), which has its salinity 31 32 33 206 maximum (± 35.76) at about 1000 m. Below, the T/S profile gradually follows the 27.75 34 35 207 kg/m3 potential density gradient towards the LSW and NADW. 36 37 38 208 39 40 209 4.1.2 Geomorphology 41 42 210 This study was carried out around a prominent spur on the La Chapelle continental 43 44 45 211 slope, flanked by deep canyons and thalweg channels in water depths from 150 to 1100 46 47 212 m (Fig. 1b). The spur has a main NE-SW orientation and an average inclination of 2°. 48 49 50 213 The slopes flanking the spur show a „herringbone‟ pattern of WNW-ESE orientated 51 52 214 gullies on the western slope (13-15°) and NNW-SSE orientated gullies on the eastern 53 54 55 215 slope (16°). 56 57 216 58 59 60 61 62 9 63 64 65

1 217 The Armorican margin near the Guilvinec Canyon is characterised by a heavily incised 2 3 218 slope with NE-SW oriented canyons and spurs (Fig. 1). The Guilvinec Canyon is 14 km 4 5 6 219 wide and bound to the northwest by the Penmarc‟h Spur. The main part of this spur is 7 8 220 relatively flat (0-2°) until 250 m water depth. Here, the gradient towards the Guilvinec 9 10 11 221 Canyon is rather abrupt (10°), especially along its relatively steep northern to north- 12 13 222 western flanks (30-40°). This flank contains about 4 large dendritic gully systems with a 14 15 16 223 NNW-SSE orientation. In contrast, the eastern flank of the canyon is much smoother 17 18 224 (and less incised) with slope gradients ranging from 5 to maximum 20°. This 19 20 21 225 asymmetry is also observed on the seismic profiles (Figs. 1c and 3). Figure 3b shows a 22 23 226 seismic profile through the largest (1 km wide) and steepest (17-35°) gully, which was 24 25 227 also examined during ROV dive B08-02 (Fig. 1c). Exclusively on the SE flank, a large 26 27 28 228 (200-400 ms TWT) sigmoidal depositional sequence can be observed which explains 29 30 229 the smoother slope texture. A comparison with other regional seismic stratigraphic 31 32 33 230 studies (Bourillet et al., 2003; Paquet et al., 2010) suggests that this sequence may be 34 35 231 correlated with the to Pleistocene Little Sole Formation. On the other hand, 36 37 38 232 almost the entire NW flank is covered with diffraction hyperbolae, suggesting erosive 39 40 233 steep slopes, irregular topography or outcropping hard substratum. Along the south- 41 42 234 western tip of the spur, six more gullies can be observed within the prolongation of the 43 44 45 235 spur‟s axis. These gullies are approximately 400-500 m wide and have a gradient 46 47 236 between 10 to 25°. 48 49 50 237 51 52 238 4.2 ROV observations 53 54 55 239 4.2.1 Dive B06-02: La Chapelle continental slope 56 57 58 59 60 61 62 10 63 64 65

1 240 Dive B06-02 followed a SW-NE, 2800-m-long track over the eastern flank of a 2 3 241 prominent spur, parallel to its elongation, and crossing several of the steep NNW-SSE 4 5 6 242 orientated gullies (Figs. 1b and 4). The main observations were carried out on the 7 8 243 central gully between 7°20‟00”W and 7°19‟40”W, allowing a more comprehensive 9 10 11 244 view on the gully environment (Fig. 4). 12 13 245 14 15 16 246 More than half (57 %) of the observed seafloor showed sandy bioturbated sediments 17 18 247 with ripple marks on the relatively flat gully shoulders (Figs. 4 and 5a). These relatively 19 20 21 248 straight NW-SE oriented sand ripples have wavelengths of approximately 10-20 cm and 22 23 249 are less than 10 cm high. On other locations outside the gully axis, a frequent (25%) 24 25 250 enigmatic pale-coloured facies was observed featuring decimetric to metric blocks or 26 27 28 251 knolls inferred to be carbonated material (Fig. 5b). No recent sediment cover was 29 30 252 observed and yellow Hexadella sp. sponges were noticed frequently on top of these 31 32 33 253 blocks. Protruding banks with a thickness ranging from 10 to 30 cm were observed 34 35 254 regularly (16%), especially on the steep slopes within the central part between 620 and 36 37 38 255 680 m (Figs. 4 and 5c-d). They form steps down to the central gully thalweg along a 39 40 256 NNE-SSW to NW-SE orientation (Fig. 4). In total, 20 laterally variable banks were 41 42 257 encountered over 60 m depth. Towards the centre of the gully they disappear into a 43 44 45 258 large N-S oriented escarpment between 630 and 650 m water depth (Fig. 5f). Based on 46 47 259 our observations, this escarpment cliff is at least 10 m high. At the base of this cliff and 48 49 50 260 the larger banks above, accumulated debris provides settling sites for sessile organisms, 51 52 261 whilst the escarpment and the larger protruding banks are sporadically colonized by 53 54 55 262 medium to dense communities of giant (10-15 cm) Neopycnodonte zibrowii oysters 56 57 263 (Figs. 5d-e), forming a 3D assemblage with occasional dead cold-water corals (Lophelia 58 59 60 61 62 11 63 64 65

1 264 pertusa). The oysters only occur in high densities (up to 100 individuals per m²) near 2 3 265 the centre of the gully between 620 and 680 m depth, but were also observed in a 4 5 6 266 similar setting at 4 other locations between 540 and 680 m (Fig. 4). Most typically, at 7 8 267 the side of the gully, they only occur underneath overhangs with a width of at least 10- 9 10 11 268 15 cm (Fig. 5d). Only nine of the observed overhangs were large enough to offer 12 13 269 sufficient space for closely-stacked, suspended shells with a medium density of about 14 15 16 270 30 individuals per m² (Fig. 5e). Near the escarpment, they form a high-density vertical 17 18 271 pavement (Fig. 5f). Many oyster individuals are still articulated and alive, or, where the 19 20 21 272 free right valve is detached, show the non-degraded and locally still dark-coloured 22 23 273 endostracum indicating that they had died only recently (Fig. 5e). 24 25 274 26 27 28 275 4.2.2 Dive B08-02: North-western flank of Guilvinec Canyon 29 30 276 Dive B08-02 provided 3700 m of seafloor observations along a U-shaped track starting 31 32 33 277 with the first stretch southwards from the NE flank of a gully at 700 m water depth and 34 35 278 ending with a second stretch parallel to the gully axis (Figs. 1c and 6). 36 37 38 279 39 40 280 Along the main part of the dive track, a featureless bioturbated sandy seafloor was 41 42 281 observed (Fig. 6). Nevertheless, the eastern part of the track revealed the presence of 43 44 45 282 abundant yellow Hexadella sp. sponges, live Madrepora oculata patches and fossil 46 47 283 Lophelia pertusa debris. Between 800 and 950 m, a rippled seabed was encountered in 48 49 50 284 10% of the frames. Especially below 900 m, straight and sometimes sinuous N-S 51 52 285 oriented low-relief sand ripples were observed with wavelengths ranging between 10-20 53 54 55 286 cm and heights of approximately 5 cm (Figs. 6 and 7c). Sporadically (less than 5% of 56 57 287 observations), the rather smoothly sloping seabed was interrupted by small banks or 58 59 60 61 62 12 63 64 65

1 288 escarpments (Figs. 6 and 7a-b). These escarpments are long “ruptures” in the seabed, 2 3 289 showing consolidated substrates and have an E-W orientation between 700 and 750 m, 4 5 6 290 while those located deeper than 750 m are usually orientated in a S-N or SSW-NNE 7 8 291 direction. Generally, the banks are up to tens of centimetres thick, while the heights of 9 10 11 292 the escarpments range between 2 to 4 m. At only three out of nine locations (Figs. 6 and 12 13 293 7a-c), the escarpments were colonized by N. zibrowii oysters and, in lesser degree, by 14 15 16 294 M. oculata. The relative abundance of N. zibrowii on these escarpments is rather low, 17 18 295 with 10 to 30 individuals per m² (Figs. 7a-b). Only on the leeward side of a 1-m-high 19 20 21 296 W-E overhanging escarpment (at 744 m depth) was a community of up to100 22 23 297 individuals per m² observed. Here, the top 40 cm underneath the edge is colonized. 24 25 298 26 27 28 299 4.2.3 Dive B08-05: Western flank of Guilvinec Canyon 29 30 300 Dive B08-05 investigated the southern shoulder of a gully south of the spur that 31 32 33 301 separates the Guilvinec from the Penmarc‟h Canyon (Figs. 1c and 8). The 3100 m long 34 35 302 track followed a southward course, later on turning to the west, covering a depth range 36 37 38 303 between 300 and 750 m. 39 40 304 41 42 305 Between 300 and 450 m, about 37% of the gently dipping slope is characterized by the 43 44 45 306 presence of straight to gently sinuous sand ripples with a wavelength between 10 to 15 46 47 307 cm and a general SSE-NNW orientation (Fig. 9a). Sometimes coarser sand was 48 49 50 308 observed in between the ripples. From 450 to 730 m, a relatively flat and bioturbated 51 52 309 silty to sandy seafloor is observed with some small escarpments near 480 m and a low- 53 54 55 310 relief rippled seabed (Fig. 8). Only at the very end of Dive B08-05, the gentle slope is 56 57 311 abruptly interrupted by a large, laterally continuous 4-m-high WSW-ENE rocky 58 59 60 61 62 13 63 64 65

1 312 escarpment at 735 m. The first meter of this escarpment is a 50 cm deep overhanging 2 3 313 cliff underneath which a thriving community of N. zibrowii oysters (Figs. 9b-c). Figure 4 5 6 314 9d shows the typical high-density assemblage of N. zibrowii, in which individuals seem 7 8 315 to have grown on top of each other. Their size reaches 10 cm on the smallest axis and 9 10 11 316 15 cm on the largest axis. The density is estimated at 63 individuals per m². Both the 12 13 317 oysters and the occasional corals are hanging upside down, concentrated near the most 14 15 16 318 overhanging end of the cliff. 17 18 319 19 20 21 320 5. Discussion 22 23 321 Compared to the previously known occurrences of Neopycnodonte zibrowii (Wisshak et 24 25 322 al., 2009a,b), these present observations in the Bay of Biscay expand the bathymetric 26 27 28 323 range of this species from 350 to 846 metres. The habitat in which these giant deep- 29 30 324 water oysters are thriving is dependent on specific conditions, both topographic and 31 32 33 325 environmental. The main part of the discussion is based on the observations made 34 35 326 during ROV dive B06-02 along the La Chapelle continental margin, since it provided 36 37 38 327 the best overview of the occurrence of N. zibrowii in relation with its environment (Fig. 39 40 328 10). 41 42 329 43 44 45 330 5.1 Influence of the physical environment on deep-water oyster colonization 46 47 331 Until now, within the Bay of Biscay, the deep-water oyster N. zibrowii has only been 48 49 50 332 observed on hard substrates. More specifically, only the areas underneath overhanging 51 52 333 cliffs, banks or steep escarpments seem to be successful colonization surfaces with 53 54 55 334 relative abundances from 30 to 100 individuals per m² (Figs. 5, 9 and 7). These specific 56 57 335 deep-water oyster substrata, which can be considered as limiting along the continental 58 59 60 61 62 14 63 64 65

1 336 margins, are concentrated within canyons, where they still are relatively sparsely 2 3 337 developed (less than 5% of the surface area in this dataset). 4 5 6 338 7 8 339 Generally, the morphology of canyons flanks is relatively irregular and no significant 9 10 11 340 cover of draping sediments is observed (Figs. 1 and 3). This zone is most likely subject 12 13 341 to reworking by along-slope (contouritic) or downslope (turbiditic) current processes 14 15 16 342 during respectively interglacial and glacial times (Arzola et al., 2008; Bourillet et al., 17 18 343 2006; Huthnance, 1995; Toucanne et al., 2009). Within the depth window 19 20 21 344 corresponding to the ROV observations, seismic profiles show single, high-amplitude 22 23 345 reflections or diffraction hyperbolae, which might correspond to lithified calcarenite or 24 25 346 calcilutites banks (Figs. 3 and 5d) from the Jones or Cockburn Formations 26 27 28 347 (Bourillet et al., 2003; Paquet et al., 2010). Within canyons and gullies, predominantly 29 30 348 downslope erosion has gradually exposed these consolidated carbonate-like sedimentary 31 32 33 349 sequences, which have been shaped into step-like banks or escarpments (Figs. 5c-d and 34 35 350 10). This process is more intense towards the centre (thalweg) of the canyon or gully, 36 37 38 351 providing a higher availability of suitable substrates for epibenthos colonization. 39 40 352 Evidently, this will also affect the nature of the sessile organisms that colonize these 41 42 353 substrates. If this epifauna is too exposed, they may be removed by episodic turbiditic 43 44 45 354 currents. The overhanging banks and the vertical escarpments may provide a 46 47 355 sufficiently sheltered habitat for the deep-water oysters, which are able to settle on such 48 49 50 356 surfaces. Along the steeper La Chapelle continental slope, suitable substrates with 51 52 357 abundant deep-water oysters were more prevalent compared to the Guilvinec area (Figs. 53 54 55 358 4 and 6). Only occasionally, and in reduced numbers, were co-occurring species such as 56 57 359 cold-water corals observed on identical substrates (Figs. 7b and 9b). Such vertical 58 59 60 61 62 15 63 64 65

1 360 substrates and overhangs could be considered as a challenging surface to be colonized 2 3 361 by cold-water corals, given their relatively weak attachment and delicate skeletal 4 5 6 362 structure. 7 8 363 9 10 11 364 The asymmetry of the Guilvinec Canyon clearly is a second factor influencing the 12 13 365 location of the deep-water oyster habitats (Fig. 3). Both morphologically and 14 15 16 366 stratigraphically the (north-)western slope of the Guilvinec Canyon shows more 17 18 367 evidence of erosion than the eastern slope, which is characterized by depositional 19 20 21 368 features (Figs. 1c and 3). Both ROV observations have shown the presence of seabed 22 23 369 ripples on different parts of this slope between 300 and 950 m (Figs. 1, 6 and 8), 24 25 370 inferring an enhanced E-W bottom current with average velocities between 20 to 40 26 27 28 371 cm/s (Stow et al., 2009). These observations fit the cyclonic flow circulation and 29 30 372 observed depth of the MOW in the Bay of Biscay (Fig. 2). The (north-)western slope of 31 32 33 373 the canyon(s) could act as an obstacle that might intensify the easterly bottom currents 34 35 374 through isopycnal doming, leading to erosion (Hernández-Molina et al., 2003; Iorga and 36 37 38 375 Lozier, 1999; Van Rooij et al., 2010). A similar process was also observed on the upper 39 40 376 part of the Portimao Canyon (Marches et al., 2007). This effect does not operate on the 41 42 377 eastern slope, leading to a preferential deposition of sediment on this side of the canyon 43 44 45 378 and making it nearly devoid of banks and escarpments (Figs. 1c and 3). This example 46 47 379 also demonstrates how a combination of turbiditic and contouritic currents may 48 49 50 380 influence canyon morphology and hence the spatial structure deep-water ecosystems. 51 52 381 Since the banks, as well as the cliffs, will not be susceptible to sediment burial, they can 53 54 55 382 provide a suitable hard substrate for this type of filter feeders. 56 57 383 58 59 60 61 62 16 63 64 65

1 384 5.2 Oceanographic drivers for deep-water oyster occurrence 2 3 385 In addition to suitable substrates on which to settle, the oysters also require a favourable 4 5 6 386 oceanographic environment and sufficient nutrients. During some of the ROV dives, 7 8 387 dense marine snow was present in the area where the oyster community was discovered 9 10 11 388 (Figs. 5a, 7c-d and 9c). This aggregated particulate matter, composed of phytodetritus 12 13 389 and faecal pellets sinking from the upper water layers, is an important constituent of the 14 15 16 390 trophic input to this community. Compared to shallow-water oysters, the occurrence of 17 18 391 oysters at these depths is remarkable given the higher salinities and generally lower 19 20 21 392 food input. On the other hand, this part of the margin is characterised by relatively high 22 23 393 surface primary production (Joint et al., 2001; Pingree and Le Cann, 1990). Moreover, 24 25 394 according to the available hydrographic data, the depth range in which N. zibrowii is 26 27 28 395 observed lies just beneath the boundary between the upper Eastern North Atlantic Water 29 30 396 and the intermediate saline Mediterranean Outflow Water (Fig. 2). The lower limit of 31 32 33 397 the oyster‟s occurrence coincides with a local salinity and temperature maximum. 34 35 398 Within this zone of upper MOW, the observation of straight sand ripples during dive 36 37 38 399 B06-02 (Figs. 4, 6 and 8) indicates the presence of strong bottom currents between 10 to 39 40 400 40 cm/s (Stow et al., 2009). Apart from the average cyclonic flow velocity of the MOW, 41 42 401 these bottom currents may be enhanced by strong internal tides (White, 2007) and 43 44 45 402 funnelled along the canyon axis, as observed along the Armorican margin (Pingree and 46 47 403 Le Cann, 1989, 1990) and within the Portimao Canyon (Marches et al., 2007). Within 48 49 50 404 the canyons, these tidally induced transports may be channelled and result in regions of 51 52 405 locally increased flow, resuspension and local circulations (Pingree and Le Cann, 1990). 53 54 55 406 Upper slope internal tides are considered to be responsible for the entrapment and 56 57 407 downward transport of enhanced fluxes of surface-derived phytodetritus (Holligan et al., 58 59 60 61 62 17 63 64 65

1 408 1985; Pingree et al., 1982), as recently inferred for the Nazaré Canyon (Arzola et al., 2 3 409 2008) and above giant carbonate mound provinces colonized by reef-forming cold- 4 5 6 410 water corals (Mienis et al., 2007; White, 2007) . The significance of submarine canyons 7 8 411 as hotspot for deep-water ecosystems may be due to their capacity to accumulate 9 10 11 412 organic debris (Canals et al., 2006; Mortensen and Buhl-Mortensen, 2005; Roberts et 12 13 413 al., 2006). It is highly likely that this water-mixing above the seabed results in enhanced 14 15 16 414 concentrations of suspended material and favors aggregations of filter/suspension 17 18 415 feeders (de Stigter et al., 2007). 19 20 21 416 22 23 417 The oyster banks occur within a potential density envelope of 27.4 – 27.7 kg/m3 (Fig. 24 25 418 2), overlapping the range of values that are considered to be a prerequisite for the 26 27 28 419 development, growth and distribution of cold-water coral reefs along the Celtic and 29 30 420 Nordic European margin (Dullo et al., 2008) and within the Guilvinec area (De Mol et 31 32 33 421 al., in press). This relationship to potential density may be linked to the formation of 34 35 422 intermediate nepheloid layers, indicating an increased nutrient supply and resuspension 36 37 38 423 (de Stigter et al., 2007; Mienis et al., 2007). Thus, in contrast to shallow marine oyster 39 40 424 occurrences, the dynamic oceanography within deep-sea canyons has the potential to 41 42 425 support a stable ecosystem in which oysters co-exist with sponges, hydrozoans, 43 44 45 426 gorgonians and scleractinians (De Mol et al., in press). 46 47 427 48 49 50 428 6. Conclusions 51 52 429 The use of ROV technology enabled detailed observations to be made of 53 54 55 430 Neopycnodonte zibrowii oysters in water depths between 540 and 846 m along the 56 57 431 French Atlantic continental margin. This recently described species can be regarded as a 58 59 60 61 62 18 63 64 65

1 432 distinct deep-sea relative of N. cochlear with specific adaptations allowing it to thrive at 2 3 433 upper bathyal depths. Although N. zibrowii was previously observed in non-canyon 4 5 6 434 environments, for example, in the Azores Archipelago (Wisshak et al., 2009a) and the 7 8 435 Gorringe Bank (Auzende et al., 1984), this species commonly requires steep slopes with 9 10 11 436 overhanging banks and escarpments. In the northern Bay of Biscay, suitable physical 12 13 437 habitats can only be generated within canyons by the interplay of turbiditic and 14 15 16 438 contouritic processes, which erode and gradually expose consolidated carbonate-like 17 18 439 sedimentary sequences, shaping them into step-like banks. Moreover, within these 19 20 21 440 canyons, the delivery and resuspension of nutrients is facilitated through funnelling of 22 23 441 internal tides near the ENAW/MOW interface. Only when these unusual physical 24 25 442 habitats and hydrographic conditions coincide can N. zibrowii be observed in relatively 26 27 28 443 high numbers of up to 100 individuals per m². 29 30 444 31 32 33 445 This little-known population located in an inaccessible environment sheds light on the 34 35 446 richness of filter-feeding species in canyon systems. Because they occupy vertical 36 37 38 447 habitats, acoustic hull-mounted systems and bottom sampling with conventional towed 39 40 448 devices are unsuitable for a proper description of these communities. Further in situ 41 42 449 observations and habitat mapping using ROVs should improve our understanding of the 43 44 45 450 biogeographic distribution of this unusual bivalved mollusc along the Bay of Biscay and 46 47 451 the European margin. It typifies an overlooked deep-water community, which represents 48 49 50 452 part of the still vastly underestimated biodiversity of bathyal benthic communities, 51 52 453 especially within canyons. 53 54 55 454 56 57 455 Acknowledgements 58 59 60 61 62 19 63 64 65

1 456 This research was supported by the HERMES project (EC contract no GOCE-CT-2005- 2 3 457 511234), funded by the European Commission‟s Sixth Framework Programme under 4 5 6 458 the priority „Sustainable Development, Global Change and Ecosystems‟ and by ESF 7 8 459 EuroDIVERSITY MiCROSYSTEMS (05_EDIV_FP083-MICROSYSTEMS). A 9 10 11 460 follow-up of this research will be performed within the framework of the EC FP7 IP 12 13 461 HERMIONE project (grant agreement n° 226354). We are indebted to the entire ROV 14 15 16 462 Genesis technical team of Ghent University (Belgium). The captains, crews and 17 18 463 shipboard scientific parties of the R/V Belgica ST0612 and ST0813a campaigns are also 19 20 21 464 acknowledged for their enthusiastic efforts. We are grateful to Dr. L. Chou (ULB, 22 23 465 Belgium) for kindly providing CTD data (Station 3 cast B). The authors also kindly 24 25 466 acknowledge H. Pirlet, three anonymous reviewers, the editor and associated editor of 26 27 28 467 DSR I for the useful comments and suggestions which significantly helped improving 29 30 468 this manuscript. LDM acknowledges the support through an IWT-grant. The research of 31 32 33 469 DVR was funded through an FWO Flanders post-doctoral fellowship. 34 35 470 36 37 38 471 Appendix A. Supplementary material 39 40 472 Supplementary data associated with this article can be found in the online version at... 41 42 473 43 44 45 474 References 46 47 475 Arzola, R.G., Wynn, R.B., Lastras, G., Masson, D.G., Weaver, P.P.E., 2008. 48 49 50 476 Sedimentary features and processes in the Nazaré and Setúbal submarine 51 52 477 canyons, west Iberian margin. Marine Geology, 250(1-2), 64-88. 53 54 55 478 Auzende, J.-M., Charvet, J., Le Lann, A., Le Pichon, X., Monteiro, J.-H., Nicolas, A., 56 57 479 Olivet, J.-L., Ribeiro, A., 1984. Géologie du Banc de Gorringe, Campagne 58 59 60 61 62 20 63 64 65

1 480 CYAGOR II. Résultats des Campagnes à la Mer 27. Publications du Centre 2 3 481 National pour l'Exploration des Oceans (CNEXO), Brest, 65 pp. 4 5 6 482 Bourillet, J.F., Reynaud, J.Y., Baltzer, A., Zaragosi, S., 2003. The 'Fleuve Manche': the 7 8 483 submarine sedimentary features from the outer shelf to the deep-sea fans. 9 10 11 484 Journal of Science, 18(3-4), 261-282. 12 13 485 Bourillet, J.F., Zaragosi, S., Mulder, T., 2006. The French Atlantic margin and deep-sea 14 15 16 486 submarine systems. Geo-Marine Letters, 26(6), 311-315. 17 18 487 Canals, M., Puig, P., de Madron, X.D., Heussner, S., Palanques, A., Fabres, J., 2006. 19 20 21 488 Flushing submarine canyons. Nature, 444(7117), 354-357. 22 23 489 Delongueville, C., Sciallet, R. 1999. Neopycnodonte cochlear (Poli, 1795) dans le Golfe 24 25 490 de Gascogne. Arion 24(2), 62 26 27 28 491 Delongueville, C., Sciallet, R. 2009. Neopycnodonte zibrowii Gofas, Salas & Taviani in 29 30 492 Wisshak et al., 2009 dans le golfe de Gascogne. Novapex 10(1), 9-12. 31 32 33 493 De Mol, L., Van Rooij, D., Pirlet, H., Greinert, J., Frank, N., Quemmerais, F., Henriet, 34 35 494 J.P., in press. Cold-water coral habitats in the Penmarc'h and Guilvinec canyons 36 37 38 495 (Bay of Biscay): Deep-water versus shallow water settings. Marine Geology. 39 40 496 de Stigter, H., Boer, W., de Jesus Mendes, P.A., Jesus, C.C., Thomsen, L., van den 41 42 497 Bergh, G.D., van Weering, T.C.E., 2007. Recent sediment transport and 43 44 45 498 deposition in the Nazaré Canyon, Portuguese continental margin. Marine 46 47 499 Geology, 246, 144-164. 48 49 50 500 Dorschel, B., Wheeler, A.J., Huvenne, V.A.I., de Haas, H., 2009. Cold-water coral 51 52 501 mounds in an erosive environmental setting: TOBI side-scan sonar data and 53 54 55 502 ROV video footage from the northwest Porcupine Bank, NE Atlantic. Marine 56 57 503 Geology, 264(3-4), 218-229. 58 59 60 61 62 21 63 64 65

1 504 Duineveld, G., Lavaleye, M., Berghuis, E., de Wilde, P., 2001. Activity and 2 3 505 composition of the benthic fauna in the Whittard Canyon and the adjacent 4 5 6 506 continental slope (NE Atlantic). Oceanologica Acta, 24(1), 69-83. 7 8 507 Dullo, C., Flögel, S., Rüggeberg, A., 2008. Cold-water coral growth in relation to the 9 10 11 508 hydrography of the Celtic and Nordic European continental margin. Marine 12 13 509 Ecology Progress Series, 371, 165-176. 14 15 16 510 Gofas, S., Freiwald, A., López Correa, M., Remia, A., Salas, C., Taviani, M., Wisshak, 17 18 511 M., Zibrowius, H., 2007. Oyster beds in the deep sea. In: K. Jordaens, N. van 19 20 21 512 Houtte, J. van Goethem, T. Backeljau (Eds), World Congress of Malacology, 22 23 513 Antwerp, Belgium, pp. 80. 24 25 514 Harry, H.W., 1981. Nominal species of living oysters proposed during the last fifty 26 27 28 515 years. Veliger, 24, 39-45. 29 30 516 Haynes, R., Barton, 1990. A poleward flow along the Atlantic coast of the Iberian 31 32 33 517 Peninsula. Journal of Geophysical Research, 95, 11425-11441. 34 35 518 Hernández-Molina, J., Llave, E., Somoza, L., Fernandez-Puga, M.C., Maestro, A., 36 37 38 519 Leon, R., Medialdea, T., Barnolas, A., Garcia, M., del Rio, V.D., Fernandez- 39 40 520 Salas, L.M., Vazquez, J.T., Lobo, F., Dias, J.M.A., Rodero, J., Gardner, J., 2003. 41 42 521 Looking for clues to paleoceanographic imprints: A diagnosis of the Gulf of 43 44 45 522 Cadiz contourite depositional systems. Geology, 31(1), 19-22. 46 47 523 Hoernle, K., Shipboard party, 2001. METEOR Cruise No. M51, Leg 1, VULKOSO - 48 49 50 524 Vulkanismus Ostatlantik-Alboran. GEOMAR, Bremen, 94 pp. 51 52 525 Holligan, P.M., Pingree, R.D., Mardell, G.T., 1985. Oceanic Solitions, Nutrient Pulses 53 54 55 526 and Phytoplankton Growth. Nature, 314(6009), 348-350. 56 57 58 59 60 61 62 22 63 64 65

1 527 Huetten, E., Greinert, J., 2008. Software controlled guidance, recording and post- 2 3 528 processing of seafloor observations by ROV and other towed devices: the 4 5 6 529 software package OFOP. Geophysical Research Abstracts, 10, EGU2008-A- 7 8 530 03088. 9 10 11 531 Huthnance, J.M., 1995. Circulation, exchange and water masses at the ocean margin: 12 13 532 the role of physical processes at the shelf edge. Progress in Oceanography, 35, 14 15 16 533 353-431. 17 18 534 Iorga, M.C., Lozier, M.S., 1999. Signatures of the Mediterranean outflow from a North 19 20 21 535 Atlantic climatology 1. Salinity and density fields. Journal of Geophysical 22 23 536 Research-Oceans, 104(C11), 25985-26009. 24 25 537 Joint, I., Wollast, R., Chou, L., Batten, S., Elskens, M., Edwards, E., Hirst, A., Burkill, 26 27 28 538 P., Groom, S., Gibb, S., Miller, A., Hydes, D., Dehairs, F., Antia, A., Barlow, 29 30 539 R., Rees, A., Pomroy, A., Brockmann, U., Cummings, D., Lampitt, R., Loijens, 31 32 33 540 M., Mantoura, F., Miller, P., Raabe, T., Alvarez-Salgado, X., Stelfox, C., 34 35 541 Woolfenden, J., 2001. Pelagic production at the Celtic Sea shelf break. Deep Sea 36 37 38 542 Research Part II: Topical Studies in Oceanography, 48(14-15), 3049-3081. 39 40 543 Le Danois, E., 1948. Les profondeurs de la mer. Payot, Paris, 303 pp. 41 42 544 Marches, E., Mulder, T., Cremer, M., Bonnel, C., Hanquiez, V., Gonthier, E., Lecroart, 43 44 45 545 P., 2007. Contourite drift construction influenced by capture of Mediterranean 46 47 546 Outflow Water deep-sea current by the Portimao submarine canyon (Gulf of 48 49 50 547 Cadiz, South Portugal). Marine Geology, 242(4), 247-260. 51 52 548 McCave, I.N., Hall, I.R., Antia, A.N., Chou, L., Dehairs, F., Lampitt, R.S., Thomsen, 53 54 55 549 L., van Weering, T.C.E., Wollast, R., 2001. Distribution, composition and flux 56 57 58 59 60 61 62 23 63 64 65

1 550 of particulate material over the European margin at 47°-50°N. Deep-Sea 2 3 551 Research II, 48, 3107-3139. 4 5 6 552 Mienis, F., de Stigter, H., White, M., Duineveld, G.C.A., de Haas, H., van Weering, T., 7 8 553 2007. Hydrodynamic controls on cold-water coral growth and carbonate-mound 9 10 11 554 development at the SW and SE Rockall Trough Margin, NE Atlantic Ocean. 12 13 555 Deep-Sea Research I, 54, 1655-1674. 14 15 16 556 Mortensen, P., Buhl-Mortensen, L., 2005. Deep-water corals and their habitats in The 17 18 557 Gully, a submarine canyon off Atlantic Canada. In: A. Freiwald, J.M. Roberts 19 20 21 558 (Eds.), Cold-Water Corals and Ecosystems. Springer-Verlag, Berlin Heidelberg, 22 23 559 pp. 247-277. 24 25 560 Pairaud, I.L., Lyard, F., Auclair, F., Letellier, T., Marsaleix, P., 2008. Dynamics of the 26 27 28 561 semi-diurnal and quarter-diurnal internal tides in the Bay of Biscay. Part 1: 29 30 562 Barotropic tides. Continental Shelf Research, 28(10-11), 1294-1315. 31 32 33 563 Palanques, A., Puig, P., Latasa, M., Scharek, R., 2009. Deep sediment transport induced 34 35 564 by storms and dense shelf-water cascading in the northwestern Mediterranean 36 37 38 565 basin. Deep-Sea Research I, 56(3), 425-434. 39 40 566 Paquet, F., Menier, D., Estournès, G., Bourillet, J.-F., Leroy, P., Guillocheau, F., 2010. 41 42 567 Buried fluvial incisions as a record of Middle-Late Miocene eustasy fall on the 43 44 45 568 Armorican Shelf (Bay of Biscay, France). Marine Geology, 268(1-4), 137-151. 46 47 569 Pingree, R.D., Le Cann, B., 1989. Celtic and Armorican slope and shelf residual 48 49 50 570 currents. Progress in Oceanography, 23, 303-338. 51 52 571 Pingree, R.D., Le Cann, B., 1990. Structure, strength and seasonality of the slope 53 54 55 572 currents in the Bay of Biscay region. Journal of the Marine Biological 56 57 573 Association of the United Kingdom, 70, 857-885. 58 59 60 61 62 24 63 64 65

1 574 Pingree, R.D., Mardell, G.T., Holligan, P.M., Griffiths, D.K., Smithers, J., 1982. Celtic 2 3 575 Sea and Armorican Current Structure and the Vertical Distributions of 4 5 6 576 Temperature and Chlorophyll. Continental Shelf Research, 1(1), 99-116. 7 8 577 Pollard, S., Griffiths, C.R., Cunningham, S.A., Read, J.F., Perez, F.F., Ríos, A.F., 1996. 9 10 11 578 Vivaldi 1991 - A study of the formation, circulation and ventilation of Eastern 12 13 579 North Atlantic Central Water. Progress in Oceanography, 37, 167-192. 14 15 16 580 Reveillaud, J., Freiwald, A., Van Rooij, D., Le Guilloux, E., Altuna, A., Foubert, A., 17 18 581 Vanreusel, A., Olu-Le Roy, K.., Henriet, J.P., 2008. The distribution of 19 20 21 582 scleractinian corals in the Bay of Biscay, NE Atlantic. Facies, 54(3), 317-331. 22 23 583 Roberts, J.M., Wheeler, A.J., Freiwald, A., 2006. Reefs of the Deep: The Biology and 24 25 584 Geology of Cold-Water Coral Ecosystems. Science, 312(5773), 543-547. 26 27 28 585 Stow, D.A.V., Hernández-Molina, F.J., Llave, E., Sayago-Gil, M., del Rio, V.D., 29 30 586 Branson, A., 2009. Bedform-velocity matrix: The estimation of bottom current 31 32 33 587 velocity from bedform observations. Geology, 37(4), 327-330. 34 35 588 Toucanne, S., Zaragosi, S., Bourillet, J.F., Cremer, M., Eynaud, F., Van Vliet-Lanoë, 36 37 38 589 B., Penaud, A., Fontanier, C., Turon, J.L., Cortijo, E., Gibbard, P.L., 2009. 39 40 590 Timing of massive 'Fleuve Manche' discharges over the last 350 kyr: insights 41 42 591 into the European ice-sheet oscillations and the European drainage network from 43 44 45 592 MIS 10 to 2. Quaternary Science Reviews, 28(13-14), 1238-1256. 46 47 593 Van Aken, H.M., 2000. The hydrography of the mid-latitude Northeast Atlantic Ocean 48 49 50 594 II: The intermediate water masses. Deep-Sea Research I, 47, 789-824. 51 52 595 Van Rooij, D., Iglesias, J., Hernández-Molina, F.J., Ercilla, G., Gomez-Ballesteros, M., 53 54 55 596 Casas, D., Llave, E., De Hauwere, A., Garcia-Gil, S., Acosta, J., Henriet, J.-P., 56 57 597 2010. The Le Danois Contourite Depositional System: Interactions between the 58 59 60 61 62 25 63 64 65

1 598 Mediterranean Outflow Water and the upper Cantabrian slope (North Iberian 2 3 599 margin). Marine Geology, 274(1-4), 1-20. 4 5 6 600 Vangriesheim, A., Khripounoff, A., 1990. Near-bottom particle concentration and flux: 7 8 601 Temporal variations observed with sediment traps and nepholometer on the 9 10 11 602 Meriadzek Terrace, Bay of Biscay. Progress In Oceanography, 24(1-4), 103-116. 12 13 603 Videt, B., Neraudeau, D., 2003. Variability and heterochronies of Rhynchostreon 14 15 16 604 suborbiculatum (Lamarck, 1801) ( : Ostreoidea : : 17 18 605 Exogyrinae) from the Cenomanian and the Lower Taronian of Charentes (SW 19 20 21 606 France). Comptes Rendus Palevol, 2(6-7), 563-576. 22 23 607 Weaver, P.P.E., Gunn, V., 2009. INTRODUCTION TO THE SPECIAL ISSUE 24 25 608 HERMES Hotspot Ecosystem Research on the Margins of European Seas. 26 27 28 609 Oceanography, 22(1), 12-15. 29 30 610 Wessel, P., Smith, W.H.F., 1991. Free Software helps Map and Display Data. EOS 31 32 33 611 Transactions AGU, 72(441), 445-446. 34 35 612 White, M., 2007. Benthic dynamics at the carbonate mound regions of the Porcupine 36 37 38 613 Sea Bight continental margin. International Journal of Earth Sciences, 96, 1-9. 39 40 614 Wisshak, M., López Correa, M., Gofas, S., Salas, C., Taviani, M., Jakobsen, J., 41 42 615 Freiwald, A., 2009a. Shell architecture, element composition, and stable isotope 43 44 45 616 signature of the giant deep-sea oyster Neopycnodonte zibrowii sp. n. from the 46 47 617 NE Atlantic. Deep Sea Research I, 56(3), 374-407. 48 49 50 618 Wisshak, M., Neumann, C., Jakobsen, J., Freiwald, A., 2009b. The `living-fossil 51 52 619 community' of the cyrtocrinid Cyathidium foresti and the deep-sea oyster 53 54 55 620 Neopycnodonte zibrowii (Azores Archipelago). Palaeogeography, 56 57 621 Palaeoclimatology, Palaeoecology, 271(1-2), 77-83. 58 59 60 61 62 26 63 64 65

1 622 Zaragosi, S., Auffret, G.A., Faugères, J.-C., Garlan, T., Pujol, C., Cortijo, E., 2000. 2 3 623 Physiography and recent sediment distribution of the Celtic Deep-Sea Fan, Bay 4 5 6 624 of Biscay. Marine Geology, 169, 207-237. 7 8 625 9 10 11 626 Figure captions 12 13 627 14 15 16 628 Figure 1: (a). Location of the study areas along the French Atlantic continental margin 17 18 629 (GEBCO bathymetry, contour lines every 250 m), with indication of the CTD locations 19 20 21 630 (Table 2). (b). Detail of the La Chapelle slope area with EM1002 bathymetry (contour 22 23 631 spacing 25 m) and the location of ROV dive B06-02 (white). (c). Detail of the Guilvinec 24 25 632 Canyon area with EM1002 bathymetry (contour spacing 25 m), with the location of 26 27 28 633 seismic profiles (red) and ROV dives B08-02 and B08-05 (white). 29 30 634 31 32 33 635 Figure 2: Hydrographic data of the two study areas. (a). Temperature/salinity plot for 34 35 636 both CTD casts (Table 2), with indication of the boundary (dashed grey line) between 36 37 38 637 the Eastern North Atlantic Water (ENAW) and the Mediterranean Outflow Water 39 40 638 (MOW). The red and blue dashed envelopes refer to the occurrence of cold-water coral 41 42 639 (CWC) reefs in respectively the Porcupine Seabight (Dullo et al., 2008) and within the 43 44 45 640 Guilvinec Canyon (De Mol et al., in press). The estimated occurrence envelope of deep- 46 47 641 water oysters in the Bay of Biscay is based on the ROV observations (Table 1), plotted 48 49 50 642 on the CTD data of, respectively, (b) cast “Station 3 cast B” and (c) cast “B0813-CTD- 51 52 643 4” in full grey. Here, the black vertical bars indicate the corresponding depth range of 53 54 55 644 ROV observations. 56 57 645 58 59 60 61 62 27 63 64 65

1 646 Figure 3: Seismic profiles Ga080605 (A) and Ga080604 (B), illustrating the seismic 2 3 647 stratigraphy and thickness of the sedimentary cover across the Guilvinec Canyon (Fig. 4 5 6 648 1c). Note the asymmetry of canyon morphology and the difference between the 7 8 649 sedimentary processes on both flanks. The white dashed lines indicate additional 9 10 11 650 unconformities within the main sequences. These figures have a 10-fold vertical 12 13 651 exaggeration. 14 15 16 652 17 18 653 Figure 4: ROV dive B06-02 track superimposed on the R/V Belgica EM1002 19 20 21 654 bathymetry (contour lines every 50 m) with indication of the recognized lithologies, 22 23 655 seabed features and the location of the imagery shown in Fig. 5. 24 25 656 26 27 28 657 Figure 5: Video-derived images of oyster assemblages and facies recognized during 29 30 658 dive B06-02 (Fig. 4). Each image bears depth information and orientation. (a) NW-SE 31 32 33 659 straight sand ripples, (b) carbonate knolls with Hexadella sp. sponge, (c) carbonate 34 35 660 banks in a series of steps, (d) overhanging carbonate bank with (low-density) oyster 36 37 38 661 community attached below the bank, in association with a dead cold-water coral, (e) 39 40 662 close-up of a probably living (high density) deep-water oyster community, (f) steep 41 42 663 oyster cliff, note the altitude of the ROV (5.3 m above sea floor). 43 44 45 664 46 47 665 Figure 6: ROV dive B08-02 track superimposed on the R/V Belgica EM1002 48 49 50 666 bathymetry (contour lines every 50 m) with indication of the recognized lithologies, 51 52 667 seabed features and the location of the stills imagery shown in Fig. 7. 53 54 55 668 56 57 58 59 60 61 62 28 63 64 65

1 669 Figure 7: Stills images of oyster assemblages and facies recognized during dive B08-02 2 3 670 (Fig. 6); (a) E-W escarpment (± 3 m high) with frequent oyster colonization, (b), S-N 4 5 6 671 oriented escarpment with a mixed solitary oyster and cold-water coral community 7 8 672 (Lophelia pertusa and the black corals Parantipathes sp., Stichopathes sp., Trissopathes 9 10 11 673 sp.), (c) sinuous S-N oriented sand ripples with sparse live solitary Madrepora oculata, 12 13 674 (d) leeward side of a W-E outcropping and overhanging rock (± 1 m high), colonized by 14 15 16 675 deep-water oysters. 17 18 676 19 20 21 677 Figure 8: ROV dive B08-05 track superimposed on the R/V Belgica EM1002 22 23 678 bathymetry (contour lines every 50 m) with indication of the recognized lithologies, 24 25 679 seabed features and the location of the stills imagery shown in Fig. 9. 26 27 28 680 29 30 681 Figure 9: Stills images of oyster assemblages and facies recognized during dive B08-05 31 32 33 682 (Fig. 8); (a) sinuous SSE-NNW oriented sand ripple field and the fish Chimera 34 35 683 monstrosa and Helicolenus dactylopterus, (b) overhanging cliff with a thriving mixed 36 37 38 684 oyster and Madrepora oculata colony, (c) distant view of a S-N colonized overhanging 39 40 685 cliff, (d) detailed view of the densely packed deep-water oyster community. 41 42 686 43 44 45 687 Figure 10: Conceptual sketch of the deep-water oyster environment, based on ROV dive 46 47 688 B06-02 on the La Chapelle continental slope. The main oyster communities are located 48 49 50 689 at the gully axis, fed by downslope currents enabling a suitable nutrient supply. The 51 52 690 banks upon which the oysters are seated probably are Miocene carbonate banks 53 54 55 691 (calcarenites). 56 57 58 59 60 61 62 29 63 64 65 Table 1 Click here to download Table: Table 1.doc

Table 1: Metadata of the ROV Genesis dive tracks

Dive Location Start End Oyster

name Coordinates Time & Coordinates Time & depth

& Date Depth Depth

B06-02 La Chapelle 47°33.34’N 08:31 47°34.07’N 15:25 540 –

17 June slope 7°20.69’W 587 m 7°19.44’W 557 m 680 m

2006

B08-02 North-western 46°56.27’N 11:24 46°55.72’N 15:46 720 –

1 June flank 5°22.89’W 712 m 5°22.90’W 900 m 846 m

2008 Guilvinec

Canyon

B08-05 Western flank 46°55.68’N 11:17 46°55.16’N 14:27 737 -

3 June Guilvinec 5°28.52’W 305 m 5°29.00’W 737 m 740 m

2008 Canyon

1 Table 2 Click here to download Table: Table 2.doc

Table 2: Metadata of the CTD casts

Name Date Location Latitude Longitude Depth

Station 3 cast B 1 June 2006 La Chapelle slope 47°25.00’N 7°16.00’ W 1396 m

B0813-CTD-4 4 June 2008 Guilvinec canyon 46°54.53’N 5°21.26’W 1404 m

1 Figure 1 Click here to download high resolution image Figure 3 Click here to download high resolution image Figure 4 Click here to download high resolution image Figure 5 Click here to download high resolution image Figure 6 Click here to download high resolution image Figure 7 Click here to download high resolution image Figure 8 Click here to download high resolution image Figure 9 Click here to download high resolution image Figure 10 Click here to download high resolution image Figure 2 Click here to download high resolution image Supplementary material Dive B06-02 Click here to download Supplementary material: B06-02.mpg Supplementary material Dive B08-02 Click here to download Supplementary material: B08-02.mpg Supplementary material Dive B08-05 Click here to download Supplementary material: B08-05.mpg